What a song does that a flashcard can't
You can still sing the words to a song you never once sat down to learn, while the vocabulary you actually studied has quietly gone. That gap isn't an accident — and the research on singing and foreign words explains it, along with the small print nobody quotes.
Somewhere in your head is a song you haven't heard in fifteen years, and you could still sing most of it right now. You never studied those words. You didn't test yourself on them, or space out your reviews, or make a single flashcard. They just moved in one summer and never left. Meanwhile the vocabulary you sat down and genuinely tried to memorise last month — the words you meant to keep — has thinned out to a handful you're no longer sure of.
It's a lopsided little injustice, and it points at something real. A melody seems to do, for free and against your will, exactly the thing effortful study struggles to do on purpose: it makes words stay. The obvious question is whether that free ride can be booked deliberately — whether setting a foreign phrase to a tune actually helps you hold it, or just feels as though it should. As it happens, people have run the experiment.
The Hungarian experiment
The clearest test comes from a 2014 study by Karen Ludke, Fernanda Ferreira and Katie Overy, published in Memory & Cognition. Sixty English-speaking adults, none of whom knew a word of Hungarian, spent fifteen minutes in a "listen-and-repeat" session with a set of Hungarian phrases. The trick was in how they repeated them. One group simply spoke the phrases back. A second spoke them back rhythmically, in time. The third sang them back, to short melodies supplied by the researchers.
Then everyone was tested — asked to recall the phrases, to produce them aloud, to show what had gone in. The singing group came out ahead, and not by a whisker: on the tests that mattered most, the ones that required actually recalling and speaking the Hungarian, their advantage over both the plain-speaking and the rhythm-only groups was statistically reliable. Crucially, the researchers checked whether this was really about music or just about the sort of person who ends up in the singing condition. It wasn't a fluke of personality: the effect held after accounting for age, gender, mood, musical training, and even each participant's raw ability to hold sounds in mind. Singing the phrase, rather than speaking it, was doing the work.
Why a tune holds words in place
The rhythm-only group is the revealing part. If melody were just a fancy way of slowing you down and marking the beat, then tapping out the rhythm should have captured most of the benefit. It didn't. So a tune is giving your memory something beyond timing.
What it gives is structure — and structure is scaffolding for recall. The psychologist Wanda Wallace laid this out in a careful 1994 paper in the Journal of Experimental Psychology, and her account still fits. A melody, she argued, does far more than carry a rhythm. It chunks a stream of words into singable phrases, fixes how long each line runs, pins down where the stresses fall, and tells you, when you're groping for the next word, roughly what shape it has to be. A tune quietly narrows the field of what could come next. The word isn't floating alone, waiting to be dredged up; it's slotted into a frame that hands you half the answer before you've consciously reached for it.
A melody isn't decoration laid over the words. It's a container built to their measure — and words carried in a container are far harder to spill than words held loose in the hand.
The catch nobody sings about
Here is where the poster-slogan version — just sing it! — falls apart, and where Wallace's work earns its keep. The mnemonic magic isn't in singing as such. It's in a tune you have genuinely got hold of.
In her experiments, melody helped recall only when the same melody repeated, verse after verse, so that the tune itself became familiar and could act as a stable frame. When people heard only a single verse, or — worse — a different melody for every verse, the music stopped helping and started hurting: they actually remembered the words better when the text was simply spoken. An unfamiliar, ever-changing tune is not a container at all. It's a second thing to learn, competing for room with the words you were trying to keep. This is exactly why the song from your teens works so well: you have sung that melody a thousand times, so it costs you nothing and holds everything.
Out of the lab, into a classroom
A fifteen-minute session with invented phrases is a clean test but a small one, so it's worth knowing the effect survives contact with real learners and real words. In a 2023 study in the journal Languages, Yuan Zhang, Florence Baills and Pilar Prieto worked with ninety-five Chinese teenagers learning English, spread across three sessions. Every learner practised the lyrics of songs whose melodies they already knew; half sang the lyrics, half recited them. Both groups improved — practice is practice — but the singers came out ahead on the measures that count for vocabulary: naming pictures, recalling what words meant, and pronouncing them. Note the setup that made it work: familiar melodies, exactly as Wallace's rule predicts.
A few honest limits
It would be easy to over-sing this. The wider literature is genuinely mixed: alongside the wins, there are careful studies that found no advantage for sung over spoken material, and some that found spoken material was recalled better once the melody was novel or the timing was matched. Singing is not a cheat code, and most of the encouraging results cluster around phrases and pronunciation and the early stages of learning — not around cramming a thousand isolated nouns before an exam.
Two practical cautions follow from all this. First, you cannot set your entire textbook to music, and you shouldn't try; the effect is a tool for the handful of stubborn phrases that refuse to stick, not a system for everything. Second — and this is the freeing part — none of it rewards a good voice. The benefit comes from the melodic structure doing its scaffolding, not from hitting the notes, so an off-key mutter under your breath is enough. The safe summary is smaller than the headline but still useful: for a phrase that won't stay, hang it on a tune you already know by heart, and let the tune remember the shape while you supply the words.
Where vocabla fits
An app can't hum for you, and vocabla doesn't pretend to. But the reason a melody works is a reason we've been building around all along: memory holds a word best when it has several different threads to pull on, not one. Singing adds the thread of a familiar tune; saying a word aloud adds the feel of your own mouth; meeting it inside a real sentence adds the grip of context. They stack. The next time a particular phrase keeps sliding off the card, try lending it a scrap of a tune you could sing in your sleep — then let vocabla's spaced returns bring it back around over the coming days, so the melody and the meaning get to reinforce each other more than once.
Most of what we write about here is quiet and invisible — chemistry, timing, the slow work of consolidation. This one you can actually hear. Take the word you keep losing, find it a tune, and sing it badly to yourself on the walk home. It's the rare study trick you'll catch yourself enjoying — and some of those words will still be there next week that otherwise wouldn't have been.